Conformance Checking of Processes Based on Monitoring Real Behavior
|
|
- Liliana Mills
- 6 years ago
- Views:
Transcription
1 Conformance Checking of Processes Based on Monitoring Real Behavior Seminar - Multimedia Retrieval and Data Mining Aljoscha Steffens Data Management and Data Exploration Group RWTH Aachen University Germany Aljoscha.Steffens@Rwth-Aachen.de Abstract. Process models are used to depict complex processes in a structured way. Processes generally consist of two different parts: the idealized model version that is often created before the execution and the execution of the process in reality. Companies nowadays possess not only the knowledge about the process model, but also about the real world execution by logging events. This poses the question if the model conforms to reality. The paper Conformance Checking of Processes Based on Monitoring Real Behavior from A. Rozinat and W.M.P. van der Aalst [1] presents ideas and metrics of how to measure this conformance. The first measure is called Fitness and shows how good the logged events fit to the underlying model, that means if they could have been generated by a process that follows the model. It is shown that this measure alone is important but not sufficient to measure the conformance so the concepts of Behavioral Appropriateness and Structural Appropriateness are introduced. They evaluate if the model represents the process in a suitable way, that means if the model represents the behavior that actually takes place and if it is minimal in structure. 1 Introduction The usage of Business Process Models in corporations increases due to the wide range of applications they can be employed for and information they offer. They can be used to provide insight into a process, to help discussing the process with stakeholders, for documentation purposes, performance analysis, to find errors in systems or procedures and much more [3]. Due to the mentioned grows in application, these models are often developed by non-expert modelers [5] or generated artificially. Algorithms for the automated generation of process models derive them with the help of event logs (see 2.1). So there is a process that is monitored on the one hand and a model that describes the process on the other hand. This poses the question if they align properly. The reasons for a misalignment are manifold but can be seen from two different perspectives: Either the model is wrong or the process wasn t executed in a proper manner. The model can be wrong for different reasons. The process itself might have
2 2 A. Steffens changed over time due to changes of laws or changes in the management. It could be that the process wasn t thoroughly planned and thus wrong from the beginning. In this cases the model can be repaired [1] or constructed anew. If the process wasn t executed correctly, the reason could be a schedule that is too tight for workers to fulfill, or that workers didn t behave how they should which could identify fraud. The reasons for both viewpoints are numerous and may not be easily visible but require the help of domain experts. The metrics and ideas that are proposed in [1] can help to indicate discrepancies between model and reality and show how it may be possible to localize them. 2 Preliminaries The following two sections explain the concepts of event logs and Petri-Nets with the help of a simplified example of a car assembly. The third section shows the basic functionality of the process discovery algorithms that were used in this paper to construct example models, namely the Alpha-Algorithm and ILP- Miner. 2.1 Event log A process generally consists of several activities, i.e. steps that have to be executed for the process to complete. A strongly simplified example of the assembly of a car is: Individual components need to be gathered, then they need to be assembled and finally the vehicle is tested. If everything works as expected, the car gets shipped, otherwise some parts need to be changed (the event log and Petri-Net that correspond to this example are depicted in figure 1). The event log captures information about this and typically consists of three layers. The first layer lists all different cases. A case contains all activities/events that are part of one distinct execution, so in the example it would consist of all activities that were executed in order to produce one car. And an activity itself combines all pieces of information that are gathered in one step of execution. The amount of information within an activity can vary a lot, generally it contains information like the task that was executed, start- and end-time, persons that were involved and the case-id it belongs to. Typically not all information is needed at the same time. So for example the automatic generation of process models from event logs just needs information about the type of task (in the following this is represented by a letter) and the order of execution. Since a lot of differentiating attributes are left out, a lot of cases become indistinguishable and are thus referred to as traces. So the condensed version of a event log just lists all different traces and shows how often they were executed. In addition to normal, that means visible tasks, a Business Process Model can also contain invisible tasks. Invisible tasks are those which don t have a corresponding log event. They can appear if, for example, the task is not observable (or logable) in the real world. An example could be a telephone call: If the caller
3 Conformance Checking 3 does not log the event him-/herself, there will be no entry in the log. In process models invisible tasks can be used for routing purposes (e.g. to bypass a visible task) or to delay visible tasks. Fig. 1: Simplified process model and event log for the assembly of a car 2.2 Petri-Net Petri-Nets are bipartite graphs that can be used to depict Business Process Models (see figure 1). They consist of four different elements: places (depicted as circles), transitions (depicted as squares), arcs (depicted as arrows) and tokens (depicted as black dots). Transitions represent the different steps that the model consists of, so they are linked to activities in the real world. The model itself is static, but tokens can flow through the network: Places can contain tokens and a transition is enabled if all preceding places contain tokens. When a transition is enabled, it may fire, whereas one token of each input place is consumed and one token on each output places is produced. The distribution of tokens represents the state that the process is in, i.e. it shows which steps can possibly be executed. A formal definition of Petri-Nets is given in [3] as following:
4 4 A. Steffens Definition 1 (Petri-Net) A Petri-Net is a triplet N = (P, T, F ) where P is a finite set of places, T is a finite set of transitions such that P T = and F (P T ) (T P ) is a set of directed arcs, called the flow relation. A marked Petri-Net is a pair (N, M) where N = (P, T, F ) is a Petri-Net and M B(P ) is a multi-set over P denoting the marking of the net. A subset of Petri-Nets are the so called Workflow-Nets. They have a dedicated start- and end-place which is convenient for describing Business Process Models. Also every node is on a path from source to think. Most of the models that are shown in this paper are Workflow-Nets. A formal definition of Workflow-Nets is given in [3] as following: Definition 2 (Workflow-Net) Let N = (P, T, F, A, l) be a (labeled) Petri-Net and t a fresh identifier not in P T. N is a Workflow-Net if and only if (a) P contains an input place i (also called source place) such that i =, (b) P contains an output place o (also called sink place) such that o = and (c) N = (P, T t, F {(o, t), ( t, i)}, A {τ}, l {( t, τ)}) is strongly connected, i.e., there is a directed path between any pair of nodes in N. 2.3 Process discovery algorithms This section gives an overview of the process discovery algorithms that were used in this paper to generate process models from logs. The goal was not to generate models that fit as good as possible to the log files. They were just used to generate a lot of different models from different logs to identify strength and weaknesses of the quality measures that are presented in [1]. α-algorithm The α-algorithm is one of the oldest algorithms that is used for process discovery. Based on four relations between tasks it generates a Workflow-Net from an event log (W ). The relations are: a>b iff there is a trace σ = t 1 t 2 t 3...t n 1 and i {1,...n 2} such that σ W and t i = a and t i+1 = b a b iff a>b and b a a#b iff a b and b a a b iff a>b and b>a The algorithm consists of 8 steps: 1. Generate the set of transitions based on the log file 2. Generate the set of start transitions (transitions that appear first in a trace) 3. Generate the set of end transitions (transitions that appear last in a trace) 4. Determine which transitions are connected and derive the possible places in the Workflow-Net (based on the relations given above) 5. Remove places that appear more than once between transitions (remove all non-maximal pairs )
5 Conformance Checking 5 6. Generate source and sink place 7. Generate arcs 8. Return Workflow-Net In the original version the α-algorithm is not able to generate loops with two elements or less and was thus improved in later versions (for example in the α + -algorithm that is part of the ProM software) [9]. ILP-Miner Another process discovery algorithm that is part of the ProM software is the ILP-Miner [6]. The ILP-Miner is an algorithm, that returns a Petri-Net which is optimal with respect to the fitness-metric. It uses Integer Linear Programming to decide whether adding a place to the model would reduce the fitness. If so, the place won t be added. A problem of this algorithm is the runtime: Since basically every combination is tested the runtime is exponential in the number of transitions. So even for an input with view activities (<20) the algorithm takes a lot longer to finish than for example the α-algorithm. 3 Quality Measures At first glance, the question of whether a model and a log conform could be answered by checking if the model is able to generate the traces in the log. This is an important property and is measured by the Fitness metric. Even though that would technically be enough, in practice the term conform implies more. It is very easy to construct simple models that are able to generate the desired log traces as seen in figure 2. The problem is that they don t describe the underlying process very well. The Flower-Model is able to generate the log traces, but also allows for other traces of unrestricted length with an arbitrary execution order of the tasks. That means it is too general. The Sequential-Model is the complete opposite: it just lists the log traces. Even if this may be all the log traces that the model should allow for, this is still not a desired structure of a model as it does not reveal any information about the corresponding process. That means it does not help to gain insight about the process nor is it of any more use than the process log itself. This shows that another dimension is needed to check whether model and log conform, which is called Appropriateness since it measures if the model describes the process appropriately. 4 Fitness As presented above, the fitness metric evaluates if the traces in a log could actually be generated by the model. A first approach could be to divide the number of possible traces by the number of all traces. This is a valid idea and can be used to measure fitness but it neglects the fact, that two traces can be false to a different degree. A trace that s completely false should be punished more
6 6 A. Steffens Fig. 2: Left: Flower-Model (black bars indicate invisible tasks), Right:Sequential- Model, both possible models for the traces <A, B, C, D, E, F>, <A, F>, <A, B, C, D, D, F>, <A, D, E, F> and <A, B, C, D, E> than a trace that s just missing a single transition. To incorporate the amount of correctness, each trace is replayed with the model. Replay means, that one tries to generate the trace with the model. That however is done in a non-blocking, log-based manner. That means every transition that should be executed will be executed. If the previous places in the Petri-Net are not occupied, or if the following places are, this defectiveness will be captured but won t obstruct the replay. This leads to four counters that capture the behavior during replay: produced tokens, consumed tokens, missing tokens and remaining tokens. If a transition can be executed without errors, i.e. all previous places are occupied and all following places are free, the counter for consumed tokens increases by the number of incoming arcs and the counter for produced tokens increases by the number of outgoing arcs. If previous places are empty, they will be filled and the counter for missing tokens increases accordingly. The counter for remaining tokens counts all tokens that are still in the Petri-Net when the replay is finished. An example for the replay of a log is given in figure 3. The process model that is shown here was automatically constructed by the Alpha-Algorithm [8] that is part of the ProM software. The traces that were used are: <A, B, D, E, F>, <A, D, B, E, G>, <A, D, C, E, H, C, D, E, H, B, D, E, F> and <A, C, D, E, F>. Some intermediate steps were left out for simplicity reasons. The following metric can be constructed from the counters to quantify the Fitness of the model.[1]. Metric 1 (Fitness) f = 1 k 2 (1 i=1 n im i k i=1 n ) + 1 k ic i 2 (1 i=1 n ir i k i=1 n ) (1) ip i In the metric k is the number of different traces, n i the amount of occurrences of the current trace, m i, r i, c i and p i the counters for missing, remaining,
7 Conformance Checking 7 Fig. 3: Example steps of the log replay with corresponding counters. The + - symbol marks places where a token was missing.
8 8 A. Steffens consumed and produced tokens. Note that there is always at least one token produced at the start-place and one token consumed at the end-place, so the denominators can t be zero. Also for all i, m i c i and r i p i so that 0 f 1. Beside a metric that allows a more sophisticated view than just counting the number of invalid traces, the replay of a log may offer meaningful insight in the shortcomings of the model. If after a replay one can observe for example that before a certain transition there is a place with a lot of remaining tokens and afterwards a place with a lot of missing ones, one can assume that there is a scenario in the real world that makes the event being omitted for a certain reason. A domain expert can now use the information to investigate further and decide for example that the deviation is correct and that the model should be adapted to allow for it. Or that the step cannot be neglected so that the execution of the process should be controlled more frequently. As for example in figure 3 there are two modifications of the model that achieve a better Fitness. In the current state, it is not possible to generate any valid log since transition E can t be executed. For E to be executed, either B or C need to be executed and for them the execution of H and thus E is necessary. It would be possible though, if there was only one place that connects A and H to both B and C. The other modification is a connection between H and the place before D. This of course is a rather simple model and thus the problems can easily be spotted. In reality those models are typically a lot bigger and more complex. In case that the model was created thoroughly at the beginning and only changed gradually over time, the structural problems should be limited to certain parts of the model and therefore be relatively easy to spot and repair after an exhaustive replay of the log. If, on the other hand, the model was generated from an incomplete log file by an algorithm and the underlying process included a lot of concurrency and loops, the resulting model can become very hard to read and even with the information about missing or remaining tokens it can be hard to locate errors and enhance the model. Other than conformance checking, a replay can be used to detect bottlenecks by extending the model with frequencies and temporal information. It can be used to construct predictive models and to offer operational support [3]. 5 Appropriateness As mentioned above, it is easy to create models that have a Fitness of one, the structural extreme examples are the Flower-Model that allows for arbitrary behavior and is thereby too general and the Sequential-Model that just lists the different log traces and is thus too specific. So it s the common discrepancy between under- and overfitting. In [1] those properties are measured separately in terms of Behavioral Appropriateness and Structural Appropriateness. 5.1 Behavoral Appropriateness Behavioral Appropriateness addresses the issue of underfitting. It measures how much behavior is allowed by the model and compares it to how much can actually
9 Conformance Checking 9 be found in the log. If the log does not include the behavior that is possible according to the model, either the log is incomplete or the model allows for behavior that is not necessary, that means it s too general. The first approach of measuring this property that is proposed in [1] also relies on replaying the log entries with the model. The idea is, that if a model is very general, it will have many alternative routes and thus many transitions will be enabled during replay (for the Flower-Model every transition will be enabled at any time). The metric is defined as follows: Metric 2 (Simple Behavioral Appropriateness) k i=1 a B = n i( T V x i ) ( T V 1) k i=1 n i (2) In the metric k is the number of different traces, n i the amount of occurrences of the current trace, x i the mean number of enabled transitions during log replay of the current trace and T V is the number of visible tasks. The problem with this metric is, that it will rank those models high that are as specific as possible which is the Sequential-Model. The author of [1] suggests another metric that is less structural dependent, but analyses how much behavior is actually needed by the model to describe the log entries. The metric uses the concepts of tasks following and preceding each other globally. That is, if there is a task x in the trace and in the same trace there is a y afterwards, y follows x. Equivalently for precede. It doesn t matter if there is another x afterwards, the condition needs to be met only once. This captures global behavioral dependencies. There are three grades though: always following/preceding, sometimes following/preceding and never following/preceding. In [1] they are defined as following: Definition 3 (Follows relations) Two activities (x, y) are in Always Follows, Never Follows or Sometimes Follows relation in the case that, if x is executed at least once, then always, never or sometimes also y is eventually executed, respectively. Definition 4 (Precedes relations) Two activities (x, y) are in Always Precedes, Never Precedes or Sometimes Precedes relation in the case that, if y is executed at least once, then always, never or sometimes also x was executed some time before, respectively. Both relations can be computed for both model and log. The idea is to measure how much more specific the log is compared to the model. The sometimesrelations play a big role in this concept, since they show concurrent or alternative behavior and thus indicate flexibility. If the model now allows two tasks to follow each other (they are in a sometimes-f ollow-relation) and in the log this relation is more strict (either always-f ollow or never-f ollow) the model s behavior is potentially to big (always considering that the log may be incomplete). If it is the other way around (meaning the model is more specific than the log) this indicates a Fitness problem.
10 10 A. Steffens In the following metric SF l and Sm F denote the sets of follow-relations with value sometimes of log and model. SP l and Sm P denote the sets of precederelations with the value sometimes. The intersection of two sets are those elements which have the value sometimes in both model and log. Metric 3 (Behavioral Appropriateness) a B = Sl F Sm F 2 S m F + Sl P Sm P 2 S m P (3) Note that both the follows- and the precedes-relation are taken into account since the relations are not symmetric and including them both makes the metric more stable 1. Note further that, for a technical reason [1], artificial transitions for Start and End were created. The metric ranges from 0 (if in both relations model and log do not conform regarding the sometimes-elements) to 1 (they share exactly the same elements with value sometimes). Figure 4 shows a model that allows concurrent execution of B and C or D. This concurrency cannot be observed in the log. The result of the metric is a B = = 5 8. Still, the model is not necessarily false. In reality the execution of a task is not only restricted by the execution of the previous ones, but can have other reasons that are for example linked to the availability of resources. Thus it may be correct, that there is a concurrent execution of tasks B and C /D even though task B is always executed first. The reason can be, that B is executed automatically and for C and D personnel is required. If the concurrency was now removed from the model it would appear as if the execution of B is mandatory for the execution of C and D which would lead to a decreased descriptiveness of the model. This of course is only a special case but it shows that it is not sufficient to rely only on the metrics - good knowledge about the process is still necessary and cannot be neglected when evaluating a process model. As with the Fitness metric, the results that are obtained from creating the f ollows- and precedes-relation can be used to increase the model quality. If the model allows that a task is in a sometimes-f ollows-relation with itself and in the log the task never follows itself, this indicates that there might be an unnecessary loop in the model. Again, the conclusions that can be drawn from this analysis are manifold and require help of a domain expert. 5.2 Structural Appropriateness For most processes there are several different models, which are able to perform the same behavior, yet their structure may differ substantially. The problem is that there are only guidelines for developing models and those can change a lot due to personal and corporate preferences. Still there are indications for a structurally bad model such as alternative duplicate tasks and redundant invisible tasks. Those structures tend to inflate the process model and should thus be 1 See [2]
11 Conformance Checking 11 Fig. 4: A simple model that contains concurrency. avoided [1]. Duplicate tasks can have two different origins. They may be needed if a task occurs in completely different context and thus needs to be separated. An example could be that you wipe the table before and after dinner. On the other hand, there are duplicate tasks which serve the structure of the process model, but could be replaced by other techniques. The Sequential-Model consists of a lot of duplicate tasks that can easily be avoided for example by splitting the model paths after the first transition. A first intuitive measure that comes to mind when examining the sequentialized model is to compare the number of transitions and places to the number of different labels. This is captured in the following metric [1]. Metric 4 (Simple Structural Appropriateness) a S = L + 2 N L is the number of different labels ( L + 2 accounts for the start and end place) and N the number of places and transitions. Places are also taken into (4)
12 12 A. Steffens account since it is possible to inflate a model unnecessarily just by adding places. The problem here is that this metric only takes the model size into account and so just allows two models with the same behavior to be compared [1]. The following metric accounts for that by incorporating alternative duplicate tasks (T AD ) and redundant invisible tasks (T RI ). T is the number of transitions in the Petri-Net. Note that duplicate tasks are defined to be visible and thus T AD + T IR < T. Metric 5 (Structural Appropriateness) a S = T ( T AD + T IR ) T (5) The difficulty here is that one needs the number of invisible and duplicate tasks to compute the metric. According to [1] the retrieval of redundant invisible tasks is fast and can be done via a structural analysis of the model. Alternative duplicate tasks can be retrieved by a state space analysis with the help of a coverability graph. However, process models can become very large, which makes the computation very expensive, since the number of paths in a model grows exponentially. Yet there are techniques available to deal with this problem such as partial order reduction and symmetry methods [1]. This metric is targeted towards identifying models that are too sequential. As an overall measure this is a good approach, but still it does not always need to be valid. In figure 5 there are two simple models that allow for the same behavior. The first one was created by hand, the second with the ILP-Miner that is part of the ProM software [6]. According to the Structural Appropriateness metric, the second model should be preferred: It does not contain any redundant invisible tasks or alternative duplicate tasks, so it results in a value of one. The upper one has a value of a S = 5 6 = Still, the upper model should be preferred due to its simple structure. When looking at the lower model, one would think at first that, because of the occurrence of loops and concurrency, a lot of different traces can be generated by it. Simplicity is, next to Replay Fitness, Precision and Generalization, one of the four quality criterions that are generally used to evaluate results of process discovery techniques [7] and it is of great importance when the model is used as a base for discussion or documentation purposes. The given example shows that it can be necessary to neglect the Structural Appropriateness for the benefits of a simpler model.
13 Conformance Checking 13 Fig. 5: Two Petri-Nets that allow for the execution of the traces <A, B, C, E> and <A, C, D, E>. The upper one results in a worse Structural Appropriateness than the lower one, but is a lot easier to read. 6 Conclusion The metrics that are presented in [1] give a good overview of the conformance of process model and log and tackle the quality dimensions Replay Fitness, Precision and Generalization. The ideas that result in the metrics are comprehensible and allow both the comparison of two models of different structure and size as well as an estimation of the absolute quality 2. There are however examples that contradict the results of the metrics, especially when the fourth quality dimension, Simplicity, is taken into account. Sometimes it supports the understandability and readability of a model, if for example duplicate tasks are allowed. A Petri-Net that is shrank to the bare minimum of transitions and places may result in an (apparent) increase of complexity. And since one of the major purposes of process models is to discuss a process, gain insight in it or document it, this should not be the goal. So together with the presented metrics it is necessary to evaluate other properties of process models: In [4] the authors suggest quality measures that come from the field of software development, in [5] the metrics are partly based on measures from graph-theory and give a good estimation of the structural quality. Also as in most topics that have more than one quality dimension, it is not enough to solely rely on the numbers of metrics; a good knowledge about the process itself also plays a fundamental role in the evaluation. Besides the ability to evaluate a model according to the Fitness measure, it was shown that the replay technique is also able to help localizing errors in the model. However, the extent to which the replay is useful for repairing a model is 2 All three metrics are between 0 and 1
14 14 A. Steffens limited by the complexity of it, which can increase very easily, especially if the model was not created by hand. The Fitness and Behavioral Appropriateness metric can be computed quickly and don t require complex algorithms and according to [1] there are also algorithms available that speed up the computation of redundant invisible tasks and alternative duplicate tasks. All in all, it can be said, that the presented metrics can give a good overview of the conformance of process models and logs but it requires the help of domain experts to evaluate the metrics and interpret the results. References 1. A. Rozinat, W.M.P. van der Aalst Conformance Checking of Processes Based on Monitoring Real Behavior, Eindhoven University of Technology, Netherlands, A. Rozinat, W.M.P. van der Aalst Conformance Testing: Measuring the Alignment Between Event Logs and Process Models, Eindhoven University of Technology, Netherlands. 3. W.M.P. van der Aalst Process Mining: Discovery, Conformance and Enhancement of Business Processes, Published by Springer, Berlin Heidelberg, Germany, I. Vanderfeesten, J. Cardoso, J. Mendling, H. Reijers, W.M.P. van der Aalst Quality Metrics for Business Process Models, Eindhoven University of Technology, Netherlands Laura Sánchez-González, Flix Garca, Jan Mendling, Francisco Ruiz, Mario Piattini Prediction of Business Process Model Quality based on Structural Metrics, University of Castilla La Mancha, Spain. 6. T. van der Wiel Process Mining using Integer Linear Programming, Eindhoven University of Technology, Netherlands, J.C.A.M. Buijs, B. F. van Dongen, and W. M.P. van der Aalst On the Role of Fitness, Precision, Generalization and Simplicity in Process Discovery, Eindhoven University of Technology, Netherlands. 8. W.M.P. van der Aalst, T. Weijters, L. Maruster Workflow Mining: Discovering Process Models from Event Logs, IEEE Transactions on Knowledge and Data Engineering, Vol. 16, No.9, A.K.A. de Medeiros, B.F. van Dongen, W.M.P. van der Aalst, A.J.M.M. Weijters Process Mining: Extending the α-algorithm to Mine Short Loops, BETA Working Paper Series, WP 113, Eindhoven University of Technology, Netherlands, 2004
Genetic Process Mining: A Basic Approach and its Challenges
Genetic Process Mining: A Basic Approach and its hallenges A.K. Alves de Medeiros, A.J.M.M. Weijters and W.M.P. van der Aalst Department of Technology Management, Eindhoven University of Technology P.O.
More informationData Streams in ProM 6: A Single-Node Architecture
Data Streams in ProM 6: A Single-Node Architecture S.J. van Zelst, A. Burattin 2, B.F. van Dongen and H.M.W. Verbeek Eindhoven University of Technology {s.j.v.zelst,b.f.v.dongen,h.m.w.verbeek}@tue.nl 2
More informationProcess Mining Discovering Workflow Models from Event-Based Data
Process Mining Discovering Workflow Models from Event-Based Data A.J.M.M. Weijters W.M.P van der Aalst Eindhoven University of Technology, P.O. Box 513, NL-5600 MB, Eindhoven, The Netherlands, +31 40 2473857/2290
More informationProM 6: The Process Mining Toolkit
ProM 6: The Process Mining Toolkit H.M.W. Verbeek, J.C.A.M. Buijs, B.F. van Dongen, W.M.P. van der Aalst Department of Mathematics and Computer Science, Eindhoven University of Technology P.O. Box 513,
More informationTowards Process Instances Building for Spaghetti Processes
Towards Process Instances Building for Spaghetti Processes Claudia Diamantini 1, Laura Genga 1, Domenico Potena 1, and Wil M.P. van der Aalst 2 1 Information Engineering Department Università Politecnica
More informationReality Mining Via Process Mining
Reality Mining Via Process Mining O. M. Hassan, M. S. Farag, M. M. MohieEl-Din Department of Mathematics, Facility of Science Al-Azhar University Cairo, Egypt {ohassan, farag.sayed, mmeldin}@azhar.edu.eg
More informationReality Mining Via Process Mining
Reality Mining Via Process Mining O. M. Hassan, M. S. Farag, and M. M. Mohie El-Din Abstract Reality mining project work on Ubiquitous Mobile Systems (UMSs) that allow for automated capturing of events.
More informationMining Process Performance from Event Logs
Mining Process Performance from Event Logs The BPI Challenge 2012 Case Study A. Adriansyah and J.C.A.M Buijs Department of Mathematics and Computer Science Eindhoven University of Technology P.O. Box 513,
More informationMining with Eve - Process Discovery and Event Structures
Mining with Eve - Process Discovery and Event Structures Robin Bergenthum, Benjamin Meis Department of Software Engineering, FernUniversität in Hagen {firstname.lastname}@fernuni-hagen.de Abstract. This
More informationDecomposed Process Mining: The ILP Case
Decomposed Process Mining: The ILP Case H.M.W. Verbeek and W.M.P. van der Aalst Department of Mathematics and Computer Science, Eindhoven University of Technology, Eindhoven, The Netherlands {h.m.w.verbeek,w.m.p.v.d.aaalst}@tue.nl
More informationDecomposed Process Mining with DivideAndConquer
Decomposed Process Mining with DivideAndConquer H.M.W. Verbeek Department of Mathematics and Computer Science, Eindhoven University of Technology, Eindhoven, The Netherlands h.m.w.verbeek@tue.nl Abstract.
More informationNon-Dominated Bi-Objective Genetic Mining Algorithm
Advances in Computational Sciences and Technology ISSN 0973-6107 Volume 10, Number 6 (2017) pp. 1607-1614 Research India Publications http://www.ripublication.com Non-Dominated Bi-Objective Genetic Mining
More informationA Simulation-Based Approach to Process Conformance
A Simulation-Based Approach to Process Conformance Pedro M. Martins IST Technical University of Lisbon Avenida Prof. Dr. Cavaco Silva 2744-016 Porto Salvo, Portugal pedro.m.martins@tagus.ist.utl.pt Abstract.
More informationQUALITY DIMENSIONS IN PROCESS DISCOVERY: THE IMPORTANCE OF FITNESS, PRECISION, GENERALIZATION AND SIMPLICITY
International Journal of Cooperative Information Systems c World Scientific Publishing Company QUALITY DIMENSIONS IN PROCESS DISCOVERY: THE IMPORTANCE OF FITNESS, PRECISION, GENERALIZATION AND SIMPLICITY
More informationWorkflow : Patterns and Specifications
Workflow : Patterns and Specifications Seminar Presentation by Ahana Pradhan Under the guidance of Prof. Rushikesh K. Joshi Department of Computer Science and Engineering Indian Institute of Technology,
More informationBPMN Miner 2.0: Discovering Hierarchical and Block-Structured BPMN Process Models
BPMN Miner 2.0: Discovering Hierarchical and Block-Structured BPMN Process Models Raffaele Conforti 1, Adriano Augusto 1, Marcello La Rosa 1, Marlon Dumas 2, and Luciano García-Bañuelos 2 1 Queensland
More informationProcess Model Discovery: A Method Based on Transition System Decomposition
Process Model Discovery: A Method Based on Transition System Decomposition Anna A. Kalenkova 1, Irina A. Lomazova 1, and Wil M.P. van der Aalst 1,2 1 National Research University Higher School of Economics
More informationTHE SELECTION OF THE ARCHITECTURE OF ELECTRONIC SERVICE CONSIDERING THE PROCESS FLOW
THE SELECTION OF THE ARCHITECTURE OF ELECTRONIC SERVICE CONSIDERING THE PROCESS FLOW PETERIS STIPRAVIETIS, MARIS ZIEMA Institute of Computer Control, Automation and Computer Engineering, Faculty of Computer
More informationAPD tool: Mining Anomalous Patterns from Event Logs
APD tool: Mining Anomalous Patterns from Event Logs Laura Genga 1, Mahdi Alizadeh 1, Domenico Potena 2, Claudia Diamantini 2, and Nicola Zannone 1 1 Eindhoven University of Technology 2 Università Politecnica
More informationProject and Production Management Prof. Arun Kanda Department of Mechanical Engineering Indian Institute of Technology, Delhi
Project and Production Management Prof. Arun Kanda Department of Mechanical Engineering Indian Institute of Technology, Delhi Lecture - 8 Consistency and Redundancy in Project networks In today s lecture
More informationA Probabilistic Approach for Process Mining in Incomplete and Noisy Logs
A Probabilistic Approach for Process Mining in Incomplete and Noisy Logs Roya Zareh Farkhady and Seyyed Hasan Aali Abstract Process mining techniques aim at automatically generating process models from
More informationOnline Conformance Checking for Petri Nets and Event Streams
Online Conformance Checking for Petri Nets and Event Streams Andrea Burattin University of Innsbruck, Austria; Technical University of Denmark, Denmark andbur@dtu.dk Abstract. Within process mining, we
More informationProcess Mining Based on Clustering: A Quest for Precision
Process Mining Based on Clustering: A Quest for Precision A.K. Alves de Medeiros 1, A. Guzzo 2, G. Greco 2, W.M.P. van der Aalst 1, A.J.M.M. Weijters 1, B. van Dongen 1, and D. Saccà 2 1 Eindhoven University
More informationDiscovering Hierarchical Process Models Using ProM
Discovering Hierarchical Process Models Using ProM R.P. Jagadeesh Chandra Bose 1,2, Eric H.M.W. Verbeek 1 and Wil M.P. van der Aalst 1 1 Department of Mathematics and Computer Science, University of Technology,
More informationScalable Process Discovery and Conformance Checking
Scalable Process Discovery and Conformance Checking Sander J.J. Leemans, Dirk Fahland, and Wil M.P. van der Aalst Eindhoven University of Technology, the Netherlands {s.j.j.leemans, d.fahland, w.m.p.v.d.aalst}@tue.nl
More informationProcess Mining Based on Clustering: A Quest for Precision
Process Mining Based on Clustering: A Quest for Precision Ana Karla Alves de Medeiros 1, Antonella Guzzo 2, Gianluigi Greco 2, WilM.P.vanderAalst 1, A.J.M.M. Weijters 1, Boudewijn F. van Dongen 1, and
More informationThe Multi-perspective Process Explorer
The Multi-perspective Process Explorer Felix Mannhardt 1,2, Massimiliano de Leoni 1, Hajo A. Reijers 3,1 1 Eindhoven University of Technology, Eindhoven, The Netherlands 2 Lexmark Enterprise Software,
More informationLattice Tutorial Version 1.0
Lattice Tutorial Version 1.0 Nenad Jovanovic Secure Systems Lab www.seclab.tuwien.ac.at enji@infosys.tuwien.ac.at November 3, 2005 1 Introduction This tutorial gives an introduction to a number of concepts
More informationThe multi-perspective process explorer
The multi-perspective process explorer Mannhardt, F.; de Leoni, M.; Reijers, H.A. Published in: Proceedings of the Demo Session of the 13th International Conference on Business Process Management (BPM
More informationChapter 2 Basic Structure of High-Dimensional Spaces
Chapter 2 Basic Structure of High-Dimensional Spaces Data is naturally represented geometrically by associating each record with a point in the space spanned by the attributes. This idea, although simple,
More informationA PROPOSAL OF USING DEVS MODEL FOR PROCESS MINING
A PROPOSAL OF USING DEVS MODEL FOR PROCESS MINING Yan Wang (a), Grégory Zacharewicz (b), David Chen (c), Mamadou Kaba Traoré (d) (a),(b),(c) IMS, University of Bordeaux, 33405 Talence Cedex, France (d)
More informationChapter 9. Software Testing
Chapter 9. Software Testing Table of Contents Objectives... 1 Introduction to software testing... 1 The testers... 2 The developers... 2 An independent testing team... 2 The customer... 2 Principles of
More informationOnline Conformance Checking for Petri Nets and Event Streams
Downloaded from orbit.dtu.dk on: Apr 30, 2018 Online Conformance Checking for Petri Nets and Event Streams Burattin, Andrea Published in: Online Proceedings of the BPM Demo Track 2017 Publication date:
More informationConformance Testing: Measuring the Fit and Appropriateness of Event Logs and Process Models
Conformance Testing: Measuring the Fit and Appropriateness of Event Logs and Process Models A. Rozinat 1,2 and W.M.P. van der Aalst 1 1 Department of Technology Management, Eindhoven University of Technology
More informationCA441 BPM - Modelling Workflow with Petri Nets. Modelling Workflow with Petri Nets. Workflow Management Issues. Workflow. Process.
Modelling Workflow with Petri Nets 1 Workflow Management Issues Georgakopoulos,Hornick, Sheth Process Workflow specification Workflow Implementation =workflow application Business Process Modelling/ Workflow
More informationDierencegraph - A ProM Plugin for Calculating and Visualizing Dierences between Processes
Dierencegraph - A ProM Plugin for Calculating and Visualizing Dierences between Processes Manuel Gall 1, Günter Wallner 2, Simone Kriglstein 3, Stefanie Rinderle-Ma 1 1 University of Vienna, Faculty of
More informationFilter Techniques for Region-Based Process Discovery
Filter Techniques for Region-Based Process Discovery S.J. van Zelst, B.F. van Dongen, W.M.P. van der Aalst Department of Mathematics and Computer Science Eindhoven University of Technology, The Netherlands
More informationProcess Mining. CS565 - Business Process & Workflow Management Systems. University of Crete, Computer Science Department
CS565 - Business Process & Workflow Management Systems Process Mining 1 Process Mining 2 Process Mining n Workflow diagnosis aims at exploiting runtime information to assist in business process re-engineering
More informationProcess Model Consistency Measurement
IOSR Journal of Computer Engineering (IOSRJCE) ISSN: 2278-0661, ISBN: 2278-8727Volume 7, Issue 6 (Nov. - Dec. 2012), PP 40-44 Process Model Consistency Measurement Sukanth Sistla CSE Department, JNTUniversity,
More informationDealing with Artifact-Centric Systems: a Process Mining Approach
Dealing with Artifact-Centric Systems: a Process Mining Approach Guangming Li and Renata Medeiros de Carvalho 2 Abstract: Process mining provides a series of techniques to analyze business processes based
More informationBidimensional Process Discovery for Mining BPMN Models
Bidimensional Process Discovery for Mining BPMN Models DeMiMoP 2014, Haifa Eindhoven Jochen De Weerdt, KU Leuven (@jochendw) Seppe vanden Broucke, KU Leuven (@macuyiko) (presenter) Filip Caron, KU Leuven
More informationChapter 3. Set Theory. 3.1 What is a Set?
Chapter 3 Set Theory 3.1 What is a Set? A set is a well-defined collection of objects called elements or members of the set. Here, well-defined means accurately and unambiguously stated or described. Any
More informationDiscovering and Navigating a Collection of Process Models using Multiple Quality Dimensions
Discovering and Navigating a Collection of Process Models using Multiple Quality Dimensions J.C.A.M. Buijs, B.F. van Dongen, and W.M.P. van der Aalst Eindhoven University of Technology, The Netherlands
More informationData- and Resource-Aware Conformance Checking of Business Processes
Data- and Resource-Aware Conformance Checking of Business Processes Massimiliano de Leoni, Wil M. P. van der Aalst, and Boudewijn F. van Dongen Eindhoven University of Technology, Eindhoven, The Netherlands
More informationMeasuring the Precision of Multi-perspective Process Models
Measuring the Precision of Multi-perspective Process Models Felix Mannhardt 1,2, Massimiliano de Leoni 1, Hajo A. Reijers 3,1, Wil M.P. van der Aalst 1 1 Eindhoven University of Technology, Eindhoven,
More informationMining CPN Models. Discovering Process Models with Data from Event Logs. A. Rozinat, R.S. Mans, and W.M.P. van der Aalst
Mining CPN Models Discovering Process Models with Data from Event Logs A. Rozinat, R.S. Mans, and W.M.P. van der Aalst Department of Technology Management, Eindhoven University of Technology P.O. Box 513,
More informationProM 4.0: Comprehensive Support for Real Process Analysis
ProM 4.0: Comprehensive Support for Real Process Analysis W.M.P. van der Aalst 1, B.F. van Dongen 1, C.W. Günther 1, R.S. Mans 1, A.K. Alves de Medeiros 1, A. Rozinat 1, V. Rubin 2,1, M. Song 1, H.M.W.
More information6. Relational Algebra (Part II)
6. Relational Algebra (Part II) 6.1. Introduction In the previous chapter, we introduced relational algebra as a fundamental model of relational database manipulation. In particular, we defined and discussed
More informationA Quality of Service Decision Model for ATM-LAN/MAN Interconnection
A Quality of Service Decision for ATM-LAN/MAN Interconnection N. Davies, P. Francis-Cobley Department of Computer Science, University of Bristol Introduction With ATM networks now coming of age, there
More informationPart II Workflow discovery algorithms
Process Mining Part II Workflow discovery algorithms Induction of Control-Flow Graphs α-algorithm Heuristic Miner Fuzzy Miner Outline Part I Introduction to Process Mining Context, motivation and goal
More informationIssues on Decentralized Consistency Checking of Multi-lateral Collaborations
Issues on Decentralized Consistency Checking of Multi-lateral Collaborations Andreas Wombacher University of Twente Enschede The Netherlands a.wombacher@utwente.nl Abstract Decentralized consistency checking
More informationTowards Robust Conformance Checking
Towards Robust onformance hecking A. Adriansyah,.F. van Dongen, and W.M.P. van der Aalst Department of Mathematics and omputer Science Eindhoven University of Technology P.O. ox 513, 5600 M Eindhoven,
More informationTowards Automated Process Modeling based on BPMN Diagram Composition
Towards Automated Process Modeling based on BPMN Diagram Composition Piotr Wiśniewski, Krzysztof Kluza and Antoni Ligęza AGH University of Science and Technology al. A. Mickiewicza 30, 30-059 Krakow, Poland
More informationSub-process discovery: Opportunities for Process Diagnostics
Sub-process discovery: Opportunities for Process Diagnostics Raykenler Yzquierdo-Herrera 1, Rogelio Silverio-Castro 1, Manuel Lazo-Cortés 1 1 Faculty 3, University of the Informatics Sciences. Habana,
More informationOn the Formal Generation of Process Redesigns
On the Formal Generation of Process Redesigns Mariska Netjes, Hajo A. Reijers, and Wil M.P. van der Aalst Eindhoven University of Technology, PO Box 513, NL-5600 MB Eindhoven, The Netherlands {m.netjes,h.a.reijers,w.m.p.v.d.aalst}@tue.nl
More informationModelling (and Analyzing) Interorganizational Communication. Jan Martijn van der Werf
Modelling (and Analyzing) Interorganizational Communication Jan Martijn van der Werf 1 2 Interaction Interaction in networks Bob Charley Alice Dave 3 Bob Can you Charley, you do Interaction in networks
More informationConformance Checking Evaluation of Process Discovery Using Modified Alpha++ Miner Algorithm
Conformance Checking Evaluation of Process Discovery Using Modified Alpha++ Miner Algorithm Yutika Amelia Effendi and Riyanarto Sarno Department of Informatics, Faculty of Information and Communication
More informationComplexity Theory. Compiled By : Hari Prasad Pokhrel Page 1 of 20. ioenotes.edu.np
Chapter 1: Introduction Introduction Purpose of the Theory of Computation: Develop formal mathematical models of computation that reflect real-world computers. Nowadays, the Theory of Computation can be
More informationDiscovering Concurrency Learning (Business) Process Models from Examples
Discovering Concurrency Learning (Business) Process Models from Examples Invited Talk CONCUR 2011, 8-9-2011, Aachen. prof.dr.ir. Wil van der Aalst www.processmining.org Business Process Management? PAGE
More informationInteraction Testing! Chapter 15!!
Interaction Testing Chapter 15 Interaction faults and failures Subtle Difficult to detect with testing Usually seen after systems have been delivered In low probability threads Occur after a long time
More information1: Introduction to Object (1)
1: Introduction to Object (1) 김동원 2003.01.20 Overview (1) The progress of abstraction Smalltalk Class & Object Interface The hidden implementation Reusing the implementation Inheritance: Reusing the interface
More informationCHAPTER 5 GENERATING TEST SCENARIOS AND TEST CASES FROM AN EVENT-FLOW MODEL
CHAPTER 5 GENERATING TEST SCENARIOS AND TEST CASES FROM AN EVENT-FLOW MODEL 5.1 INTRODUCTION The survey presented in Chapter 1 has shown that Model based testing approach for automatic generation of test
More informationImproving Process Model Precision by Loop Unrolling
Improving Process Model Precision by Loop Unrolling David Sánchez-Charles 1, Marc Solé 1, Josep Carmona 2, Victor Muntés-Mulero 1 1 CA Strategic Research Labs, CA Technologies, Spain David.Sanchez,Marc.SoleSimo,Victor.Muntes@ca.com
More informationInteraction Testing. Chapter 15
Interaction Testing Chapter 15 Interaction faults and failures Subtle Difficult to detect with testing Usually seen after systems have been delivered In low probability threads Occur after a long time
More informationFrom Task Graphs to Petri Nets
From Task Graphs to Petri Nets Anthony Spiteri Staines Department of Computer Inf. Systems, Faculty of ICT, University of Malta Abstract This paper describes the similarities between task graphs and Petri
More informationAdministrivia. ECE/CS 5780/6780: Embedded System Design. Acknowledgements. What is verification?
Administrivia ECE/CS 5780/6780: Embedded System Design Scott R. Little Lab 8 status report. Set SCIBD = 52; (The Mclk rate is 16 MHz.) Lecture 18: Introduction to Hardware Verification Scott R. Little
More informationWelfare Navigation Using Genetic Algorithm
Welfare Navigation Using Genetic Algorithm David Erukhimovich and Yoel Zeldes Hebrew University of Jerusalem AI course final project Abstract Using standard navigation algorithms and applications (such
More informationCS/ECE 5780/6780: Embedded System Design
CS/ECE 5780/6780: Embedded System Design John Regehr Lecture 18: Introduction to Verification What is verification? Verification: A process that determines if the design conforms to the specification.
More informationTypes of Process Mining
1 Types of Process Mining 2 Types of Mining Algorithms 3 Types of Mining Algorithms 4 Control-Flow Mining 1. Start 2. 1. Get Ready Start 1. 3. 2. Start Travel by Get Train Ready 1. 2. 4. Start 3. Get Beta
More informationCore Membership Computation for Succinct Representations of Coalitional Games
Core Membership Computation for Succinct Representations of Coalitional Games Xi Alice Gao May 11, 2009 Abstract In this paper, I compare and contrast two formal results on the computational complexity
More informationIJREAT International Journal of Research in Engineering & Advanced Technology, Volume 1, Issue 5, Oct-Nov, ISSN:
IJREAT International Journal of Research in Engineering & Advanced Technology, Volume 1, Issue 5, Oct-Nov, 20131 Improve Search Engine Relevance with Filter session Addlin Shinney R 1, Saravana Kumar T
More informationUsing Minimum Description Length for Process Mining
Using Minimum Description Length for Process Mining T. alders TU indhoven t.calders@tue.nl.w. Günther TU indhoven c.w.gunther@tue.nl M. Pechenizkiy TU indhoven m.pechenizkiy@tue.nl. Rozinat TU indhoven
More informationClassification with Diffuse or Incomplete Information
Classification with Diffuse or Incomplete Information AMAURY CABALLERO, KANG YEN Florida International University Abstract. In many different fields like finance, business, pattern recognition, communication
More information2015 International Conference on Computer, Control, Informatics and Its Applications
2015 International Conference on Computer, Control, Informatics and Its Applications Business Process Optimization from Single Timestamp Event Log Riyanarto Sarno *, Fitrianing Haryadita, Kartini, Sarwosri
More informationDiscovering Hierarchical Process Models Using ProM
Discovering Hierarchical Process Models Using ProM R.P. Jagadeesh Chandra Bose 1,2, Eric H.M.W. Verbeek 1 and Wil M.P. van der Aalst 1 1 Department of Mathematics and Computer Science, University of Technology,
More informationMulti-relational Decision Tree Induction
Multi-relational Decision Tree Induction Arno J. Knobbe 1,2, Arno Siebes 2, Daniël van der Wallen 1 1 Syllogic B.V., Hoefseweg 1, 3821 AE, Amersfoort, The Netherlands, {a.knobbe, d.van.der.wallen}@syllogic.com
More informationInteroperability in the ProM Framework
Interoperability in the ProM Framework H.M.W. Verbeek 1, B.F. van Dongen 1, J. Mendling 2, and W.M.P. van der Aalst 1 1 Department of Technology Management, Eindhoven University of Technology P.O. Box
More informationANALYZING PROCESS MODELS USING GRAPH REDUCTION TECHNIQUES
NLYZING PROCESS MODELS USING GRPH REDUCTION TECHNIQUES WSIM SDIQ ND MRI E. ORLOWSK Distributed Systems Technology Centre Department of Computer Science & Electrical Engineering The University of Queensland,
More informationIntra- and Inter-Organizational Process Mining: Discovering Processes Within and Between Organizations
Intra- and Inter-Organizational Process Mining: Discovering Processes Within and Between Organizations Wil M.P. van der Aalst Eindhoven University of Technology, PO Box 513, NL-5600 MB, Eindhoven, The
More informationA Study of Quality and Accuracy Tradeoffs in Process Mining
A Study of Quality and Accuracy Tradeoffs in Process Mining Zan Huang and Akhil Kumar Smeal College of Business, Pennsylvania State University, University Park, PA 16802, USA, {zanhuang,akhilkumar}@psu.edu
More informationBehavioral Conformance of Artifact-Centric Process Models
Behavioral Conformance of Artifact-Centric Process Models Dirk Fahland, Massimiliano de Leoni, Boudewijn F. van Dongen, and Wil M.P. van der Aalst Eindhoven University of Technology, The Netherlands (d.fahland
More informationProcess Mining Put Into Context
Process Mining Put Into Context Wil M.P. van der Aalst 1,2 and Schahram Dustdar 3 1 Eindhoven University of Technology 2 Queensland University of Technology 3 Technical University of Vienna Abstract. Process
More informationAbout the Tutorial. Audience. Prerequisites. Copyright & Disclaimer. Compiler Design
i About the Tutorial A compiler translates the codes written in one language to some other language without changing the meaning of the program. It is also expected that a compiler should make the target
More informationIntroduction to Modeling. Lecture Overview
Lecture Overview What is a Model? Uses of Modeling The Modeling Process Pose the Question Define the Abstractions Create the Model Analyze the Data Model Representations * Queuing Models * Petri Nets *
More informationProcess Modelling using Petri Nets
Process Modelling using Petri Nets Katalina Grigorova Abstract: This paper discusses the reasons, which impose Petri nets as a conceptual standard for modelling and analysis of workflow. Petri nets notation
More informationPublished in: Petri Nets and Other Models of Concurrency - ICATPN 2007 (28th International Conference, Siedcle, Poland, June 25-29, 2007)
ProM 4.0 : comprehensive support for real process analysis van der Aalst, W.M.P.; van Dongen, B.F.; Günther, C.W.; Mans, R.S.; Alves De Medeiros, A.K.; Rozinat, A.; Rubin, V.A.; Song, M.S.; Verbeek, H.M.W.;
More information(Refer Slide Time: 00:01:30)
Digital Circuits and Systems Prof. S. Srinivasan Department of Electrical Engineering Indian Institute of Technology, Madras Lecture - 32 Design using Programmable Logic Devices (Refer Slide Time: 00:01:30)
More informationDiagnostic Information for Control-Flow Analysis of Workflow Graphs (aka Free-Choice Workflow Nets)
Diagnostic Information for Control-Flow Analysis of Workflow Graphs (aka Free-Choice Workflow Nets) Cédric Favre(1,2), Hagen Völzer(1), Peter Müller(2) (1) IBM Research - Zurich (2) ETH Zurich 1 Outline
More informationSoftware Design and Analysis for Engineers
Software Design and Analysis for Engineers by Dr. Lesley Shannon Email: lshannon@ensc.sfu.ca Course Website: http://www.ensc.sfu.ca/~lshannon/courses/ensc251 Simon Fraser University Slide Set: 9 Date:
More informationModel Repair Aligning Process Models to Reality
Model Repair Aligning Process Models to Reality Dirk Fahland, Wil M.P. van der Aalst Eindhoven University of Technology, The Netherlands Abstract Process mining techniques relate observed behavior (i.e.,
More informationDiscovering workflow nets using integer linear programming
Computing DOI 10.1007/s00607-017-0582-5 Discovering workflow nets using integer linear programming S. J. van Zelst 1 B. F. van Dongen 1 W. M. P. van der Aalst 1 H. M. W. Verbeek 1 Received: 19 July 2016
More informationSoftware Design Fundamentals. CSCE Lecture 11-09/27/2016
Software Design Fundamentals CSCE 740 - Lecture 11-09/27/2016 Today s Goals Define design Introduce the design process Overview of design criteria What results in a good design? Gregory Gay CSCE 740 -
More informationProcess Mining: Using CPN Tools to Create Test Logs for Mining Algorithms
Process Mining: Using CPN Tools to Create Test Logs for Mining Algorithms A.K. Alves de Medeiros and C.W. Günther Department of Technology Management, Eindhoven University of Technology P.O. Box 513, NL-5600
More informationFinding Suitable Activity Clusters for Decomposed Process Discovery
Finding Suitable Activity Clusters for Decomposed Process Discovery Details omitted for double-blind reviewing Institute omitted for double-blind reviewing Abstract. Event data can be found in any information
More informationExploring Econometric Model Selection Using Sensitivity Analysis
Exploring Econometric Model Selection Using Sensitivity Analysis William Becker Paolo Paruolo Andrea Saltelli Nice, 2 nd July 2013 Outline What is the problem we are addressing? Past approaches Hoover
More informationBusiness Process Modeling. Version 25/10/2012
Business Process Modeling Version 25/10/2012 Maurizio Morisio, Marco Torchiano, 2012, 2013 3 BP Aspects Process flow Process modeling UML Activity Diagrams BPMN Information Conceptual modeling UML Class
More information3.4 Deduction and Evaluation: Tools Conditional-Equational Logic
3.4 Deduction and Evaluation: Tools 3.4.1 Conditional-Equational Logic The general definition of a formal specification from above was based on the existence of a precisely defined semantics for the syntax
More informationDarshan Institute of Engineering & Technology Unit : 9
1) Explain software testing strategy for conventional software architecture. Draw the spiral diagram showing testing strategies with phases of software development. Software Testing: Once source code has
More informationFinding Suitable Activity Clusters for Decomposed Process Discovery
Finding Suitable Activity Clusters for Decomposed Process Discovery B.F.A. Hompes, H.M.W. Verbeek, and W.M.P. van der Aalst Department of Mathematics and Computer Science Eindhoven University of Technology,
More informationModeling Routing Constructs to Represent Distributed Workflow Processes Using Extended Petri Nets
Modeling Routing Constructs to Represent Distributed Workflow Processes Using Extended Petri Nets Mehmet Karay * Final International University, Business Administrative, Toroslar Avenue, No:6, 99370, Catalkoy,
More information